Midjourney's Version 7, released on April 3, 2025, and established as the default model on June 17, 2025, represents a significant advancement in AI-powered image generation. This model excels at interpreting both text and image prompts with exceptional precision, producing images with richer textures, more coherent details in complex elements such as human bodies, hands, and objects, and an overall enhanced aesthetic quality compared to its predecessors.

Key features of Version 7 include:

• Draft Mode: A new feature that allows users to generate images more quickly, ideal for rapid prototyping and iterative creative processes.
• Omni Reference: A powerful tool that enables users to reference multiple aspects of an image—such as its structure, style, and content—when generating new images, offering greater control over the output.
• Flexible Parameters: Version 7 supports images in any aspect ratio, strong and subtle variations, a remix feature for tweaking existing outputs, and personalization options through ranking or moodboard creation.

Compared to Version 6, released in December 2023, Version 7 discontinues some features like Character Reference and Multi-Prompting. However, it introduces improvements such as enhanced upscalers, new quality settings (allowing values of 1, 2, or 4 with distinct behaviors), and experimental parameters like --exp for advanced users. For those focused on anime-style art, Midjourney offers Niji 6, released in June 2024, as a specialized model that excels in Eastern aesthetic styles and text rendering, making it a strong alternative for specific use cases.

Midjourney is actively testing Version 7.1, which promises modest but meaningful improvements over Version 7.0. This update focuses on increasing the accuracy of prompt interpretation, refining the Omni Reference feature for more precise image referencing, and enhancing Draft Mode for even faster prototyping. The release of Version 7.1 is expected in the near future, after which development efforts will shift toward Version 8.

Version 8 is currently in active development and represents a major overhaul with a new underlying architecture and upgraded server infrastructure. It is being trained on a significantly larger dataset to expand its knowledge base and improve the intelligence of its outputs, enabling it to handle a broader range of topics and creative scenarios. While no specific release date has been confirmed, Version 8 is targeted for launch by the end of 2025. Following its release, Midjourney plans to introduce a new image editor and an advanced Version 2 video model to further expand its creative tools.

Community discussions have highlighted potential features like "Character Lock," which could allow users to maintain consistent character biometrics across different scenes, styles, or environments. If implemented, this feature would be transformative for storytelling applications such as comics, films, games, and marketing campaigns. Additionally, Midjourney is improving its video generation models, focusing on faster processing, higher quality outputs, and reduced costs as they transition from research to production. Development of 3D models has faced delays due to team health challenges but is nearing completion.

Midjourney is introducing several transformative features, many of which were highlighted during the October 16, 2025, Office Hours session:

• Stand-Alone Image Editor: Set to launch soon, this editor will allow users to edit external images (e.g., from the web or personal cameras) using Midjourney's models. Initially, it may be exclusive to Pro or Pro Yearly subscribers to ensure robust moderation.
• Depth Reference: A new tool within the image editor that enables users to re-texture images while preserving their shapes, functioning like an advanced filter for creative transformations.
• Multiple Personalization Profiles: Users will be able to create fast, tailored profiles based on demographics or preferred styles, with real-time model training integrated into the Create feature for personalized outputs.
• Storytelling Tools: Midjourney is rolling out initial storytelling tools, including community videos and talks, with plans to make these accessible to all users for applications in narrative-driven projects.
• Midjourney TV: A new feature offering grid views for videos, supporting layouts like 2x2, 3x3, or 4x4 (up to 16 videos simultaneously) at www.midjourney.tv/[number]. Users can right-click to copy the current video frame, though large grids may cause browser performance issues.
• User Profiles and Explore Enhancements: Upcoming user profiles will include follow functionality designed to minimize negativity, alongside a tenfold boost in personalization for the Explore page to better match user preferences.
• Question Submission System: A new system allows users to submit and vote on questions for Office Hours, ensuring community priorities are addressed.

Sailboat Trips (October 17): Starting October 17, 2025, Midjourney is hosting sunset catamaran cruises for community members over three consecutive weekends. These events feature music and opportunities to meet the engineering team. Interested users can sign up at midjourney.typeform.com/to/oo6kvC4Q.

Research and Interactive Content: Midjourney plans to release posts on topics like demographics, psychometrics, and collective intelligence, accompanied by standalone interactive tools to engage the community. A community experiment is also underway, allowing users to submit and vote on world problems to foster collaborative ideation.

Secret Projects: Several undisclosed initiatives, including some related to hardware, are slated for reveal over the next six months. These projects vary in scope to appeal to a broad audience, adding to the anticipation of Midjourney's future offerings.

As Midjourney introduces powerful new tools like the stand-alone image editor, the platform is intensifying its moderation efforts. This includes manual reviews of certain outputs and higher processing costs for specific jobs to ensure responsible use. Attempts to create deepfakes will result in user bans, and stricter enforcement may apply to yearly membership holders to maintain ethical standards.

The Midjourney community is buzzing with excitement on platforms like X, particularly about integrations that combine Midjourney images with tools like Grok Imagine to create quick, high-quality videos. Some third-party platforms are also beginning to bundle access to Midjourney Version 7, expanding its reach. However, the core focus remains on official updates and features from Midjourney's team, which continue to drive innovation in the AI art space.